Search NASA⌕ Search

SEARCH · Search NASA

Results for “Protein distance prediction”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

36 records · Page 2
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Accurate prediction of protein structures and interactions using a three-track neural network

DeepMind presented notably accurate predictions at the recent 14th Critical Assessment of Structure Prediction (CASP14) conference. We explored network architectures that incorporate related ideas and obtained the best performance with a three-track network in which information at the one-dimensional (1D) sequence level, the 2D distance map level, and the 3D coordinate level is successively transformed and integrated. The three-track network produces structure predictions with accuracies approaching those of DeepMind in CASP14, enables the rapid solution of challenging x-ray crystallography and cryo–electron microscopy structure modeling problems, and provides insights into the functions of proteins of currently unknown structure. The network also enables rapid generation of accurate protein-protein complex models from sequence information alone, short-circuiting traditional approaches that require modeling of individual subunits followed by docking. Here, we make the method available to the scientific community to speed biological research.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2020

Binding Affinity Prediction by Pairwise Function Based on Neural Network

In this paper, we present a new approach to estimate the binding affinity from given three-dimensional poses of protein–ligand complexes. In this scheme, every protein–ligand atom pair makes an additive free-energy contribution. The sum of these pairwise contributions then gives the total binding free energy or the logarithm of the dissociation constant. The pairwise contribution is calculated by a function implemented via a neural network that takes the properties of the two atoms and their distance as input. The pairwise function is trained using a portion of the PDBbind 2018 data set. The model achieves good accuracy for affinity predictions when evaluated with PDBbind 2018 and with the CASF-2016 benchmark, comparing favorably to many scoring functions such as that of AutoDock Vina. The framework here may be extended to incorporate other factors to further improve its accuracy and power.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Allosteric prediction via convolutional neural networks and protein structural and dynamical features

Allostery is the phenomenon whereby a binding event or covalent modification at one site in a protein modulates function at a distal site, thus changing a protein’s functional state. As such, it is a ubiquitous aspect of protein functional regulation. Computationally predicting allosteric states is important as part of the broader challenge of functional annotation, but it also has practical implications for drug development, as targeting an allosteric site often affords greater specificity compared with targeting an orthosteric site. This study introduces a machine learning approach to predict the allosteric functional state using the small G-protein KRas as the model system, due to its implication in many types of cancer and being well studied as a result with many x-ray crystallographic structures of KRas available with different mutations and ligands bound. Using structural and dynamical features that can be cast as images, namely interatomic distances, contact maps, covariance, and mutual information, supervised learning was performed using convolutional neural networks. Two pretrained convolutional neural network architectures, GoogLeNet and ResNet18, were fine-tuned to classify KRas into active or inactive states based on these features. Across training regimes, atomic contact maps emerged as the most effective structural feature, whereas linearized mutual information outperformed covariance in capturing dynamical correlations relevant to allostery. Models achieved significant validation accuracy, with atomic contact maps yielding up to 90% accuracy. In conclusion, the findings suggest that integrating global structural rearrangements and correlated motion patterns with deep learning can reliably predict protein allosteric states, offering a promising framework for understanding allosteric regulation and developing targeted therapeutics.

Rajeshwar T., Rajitha [Oak Ridge National Laborato↗
DOE OSTI2022

Spatial turnover of soil viral populations and genotypes overlain by cohesive responses to moisture in grasslands

Viruses shape microbial communities, food web dynamics, and carbon and nutrient cycling in diverse ecosystems. However, little is known about the patterns and drivers of viral community composition, particularly in soil, precluding a predictive understanding of viral impacts on terrestrial habitats. To investigate soil viral community assembly processes, here we analyzed 43 soil viromes from a rainfall manipulation experiment in a Mediterranean grassland in California. We identified 5,315 viral populations (viral operational taxonomic units [vOTUs] with a representative sequence ≥10 kbp) and found that viral community composition exhibited a highly significant distance–decay relationship within the 200-m 2 field site. This pattern was recapitulated by the intrapopulation microheterogeneity trends of prevalent vOTUs (detected in ≥90% of the viromes), which tended to exhibit negative correlations between spatial distance and the genomic similarity of their predominant allelic variants. Although significant spatial structuring was also observed in the bacterial and archaeal communities, the signal was dampened relative to the viromes, suggesting differences in local assembly drivers for viruses and prokaryotes and/or differences in the temporal scales captured by viromes and total DNA. Despite the overwhelming spatial signal, evidence for environmental filtering was revealed in a protein-sharing network analysis, wherein a group of related vOTUs predicted to infect actinobacteria was shown to be significantly enriched in low-moisture samples distributed throughout the field. Overall, our results indicate a highly diverse, dynamic, active, and spatially structured soil virosphere capable of rapid responses to changing environmental conditions.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

PNNL-Predictive-Phenomics/ProCaliper

ProCaliper is a Python library that curates, organizes, and computes protein structure features in a way that easily interfaces with user-provided experimental data. It extracts or computes protein binding site, active site, charge, pLDDT (order/disorder), acid dissociation, protonation, solvent accessible surface area, disulfide bond distance, and protein secondary structure data using precomputed protein structures and publicly available databases. It provides a unified API for integrating additional residue-level data and for visualizing residue features in 3D.

Rozum, Jordan [Pacific Northwest National Lab]↗
DOE OSTI2022

Microbial communities of Auka hydrothermal sediments shed light on vent biogeography and the evolutionary history of thermophily

Abstract Hydrothermal vents have been key to our understanding of the limits of life, and the metabolic and phylogenetic diversity of thermophilic organisms. Here we used environmental metagenomics combined with analysis of physicochemical data and 16S rRNA gene amplicons to characterize the sediment-hosted microorganisms at the recently discovered Auka vents in the Gulf of California. We recovered 325 metagenome assembled genomes (MAGs) representing 54 phyla, over 30% of those currently known, showing the microbial community in Auka hydrothermal sediments is highly diverse. 16S rRNA gene amplicon screening of 224 sediment samples across the vent field indicates that the MAGs retrieved from a single site are representative of the microbial community in the vent field sediments. Metabolic reconstruction of a vent-specific, deeply branching clade within the Desulfobacterota suggests these organisms metabolize sulfur using novel octaheme cytochrome-c proteins related to hydroxylamine oxidoreductase. Community-wide comparison between Auka MAGs and MAGs from Guaymas Basin revealed a remarkable 20% species-level overlap, suggestive of long-distance species transfer over 400 km and subsequent sediment colonization. Optimal growth temperature prediction on the Auka MAGs, and thousands of reference genomes, shows that thermophily is a trait that has evolved frequently. Taken together, our Auka vent field results offer new perspectives on our understanding of hydrothermal vent microbiology.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

OpenMDlr Dataset

Protein structure dataset created with OpenMDlr. Restraints applied to fold proteins into native-like conformations were obtained from a variety of methods; results from these different restraint sets are contained in respective subdirectories. Supplements OpenMDlr: Parallel, open-source tools for general protein structure modeling and refinement from pairwise distances (https://doi.org/10.1093/bioinformatics/btac307).

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

Protein model quality assessment using rotation–equivariant transformations on point clouds

Machine learning research concerning protein structure has seen a surge in popularity over the last years with promising advances for basic science and drug discovery. Working with macromolecular structure in a machine learning context requires an adequate numerical representation, and researchers have extensively studied representations such as graphs, discretized 3D grids, and distance maps. As part of CASP14, we explored a new and conceptually simple representation in a blind experiment: atoms as points in 3D, each with associated features. These features—initially just the basic element type of each atom—are updated through a series of neural network layers featuring rotation-equivariant convolutions. Starting from all atoms, we further aggregate information at the level of alpha carbons before making a prediction at the level of the entire protein structure. We find that this approach yields competitive results in protein model quality assessment despite its simplicity and despite the fact that it incorporates minimal prior information and is trained on relatively little data. As a result, its performance and generality are particularly noteworthy in an era where highly complex, customized machine learning methods such as AlphaFold 2 have come to dominate protein structure prediction.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Transient local secondary structure in the intrinsically disordered C‐term of the Albino3 insertase

Albino3 (Alb3) is an integral membrane protein fundamental to the targeting and insertion of light‐harvesting complex (LHC) proteins into the thylakoid membrane. Alb3 contains a stroma‐exposed C‐terminus (Alb3‐Cterm) that is responsible for binding the LHC‐loaded transit complex before LHC membrane insertion. Alb3‐Cterm has been reported to be intrinsically disordered, but precise mechanistic details underlying how it recognizes and binds to the transit complex are lacking, and the functional roles of its four different motifs have been debated. Using a novel combination of experimental and computational techniques such as single‐molecule fluorescence resonance energy transfer, circular dichroism with deconvolution analysis, site‐directed mutagenesis, trypsin digestion assays, and all‐atom molecular dynamics simulations in conjunction with enhanced sampling techniques, we show that Alb3‐Cterm contains transient secondary structure in motifs I and II. The excellent agreement between the experimental and computational data provides a quantitatively consistent picture and allows us to identify a heterogeneous structural ensemble that highlights the local and transient nature of the secondary structure. This structural ensemble was used to predict both the inter‐residue distance distributions of single molecules and the apparent unfolding free energy of the transient secondary structure, which were both in excellent agreement with those determined experimentally. We hypothesize that this transient local secondary structure may play an important role in the recognition of Alb3‐Cterm for the LHC‐loaded transit complex, and these results should provide a framework to better understand protein targeting by the Alb3‐Oxa1‐YidC family of insertases.

Okoto, Patience S.↗
DOE OSTI2023

REC protein family expansion by the emergence of a new signaling pathway

This report presents multi-genome evidence that REC protein family expansion occurs when the emergence of new pathways gives rise to functional discordance. Specificity between residues in REC domain containing response regulators with paired histidine kinases is under negative purifying selection, constrained by the presence of other bacterial two-component systems signaling cascades that share sequence and structural identity. Presuming that the two-component systems can evolve by neutral amino acid changes (neutral drift) when purifying evolutionary constraints are relaxed, how might the REC protein family expand by amino acid changes when these constraints remain intact? Using an unsupervised machine learning approach to observe the sequence landscape of REC domains across long phylogenetic distances, we find that within-gene recombination, a subcategory of gene conversion, switched the effector domain and, consequently, the regulatory context of a duplicated response regulator from transcriptional regulation by σ54 to that by σ70. We determined that the recombined response regulator diverged from its parent by episodic diversifying selection and neutral drift. Functional experiments of the parent of recombined response regulators in a model Pseudomonas putida KT2440 model system revealed that the parent and recombined response regulators sense and respond to different carboxylic acids. Finally, a residue-switching experiment using structural predictions and functional characterization suggests that the new residues in the recombined regulator could form a new interaction interface and mediate condition-specific phosphotransfer. Overall, our study finds that genetic perturbations can create conditions of functional discordance, whereby the REC protein family can evolve by episodic diversifying selection.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Light-induced electron spin qubit coherences in the purple bacteria reaction center protein

Photosynthetic reaction center proteins (RCs) provide ideal model systems for studying quantum entanglement between multiple spins, a quantum mechanical phenomenon wherein the properties of the entangled particles become inherently correlated. Following light-generated sequential electron transfer, RCs generate spin-correlated radical pairs (SCRPs), also referred to as entangled spin qubit (radical) pairs (SQPs). Understanding and controlling coherence mechanisms in SCRP/SQPs is important for realizing practical uses of electron spin qubits in quantum sensing applications. The bacterial RC (bRC) provides an experimental system for exploring quantum effects in the SCRP P 865 + Q A − , where P 865 , a special pair of bacteriochlorophylls, is the primary donor, and Q A is the primary quinone acceptor. In this study, we focus on understanding how local molecular environments and isotopic substitution, particularly deuteration, influence spin coherence times (T M ). Using high-frequency electron paramagnetic resonance (EPR) spectroscopy, we observed that the local environment surrounding P 865 and Q A plays a significant role in determining T M . Our findings show that while deuteration led to a modest increase in T M , particularly at low temperatures, but the effect was substantially smaller than predicted by classical nuclear spin diffusion alone. This result is in contrast to our previous study of the photosystem I (PSI) RC, where no increase in T M was observed upon deuteration. Theoretical modeling identified several methyl groups at key distances from the spin centers of both bRC and PSI, and methyl group tunneling at low temperatures has been previously suggested as a mechanism for enhanced spin decoherence. Additionally, our study revealed a strong dependence of spin coherence on the orientation of the external magnetic field, highlighting the influence of the protein microenvironment on spin dynamics. In conclusion, these results offer new insights for optimizing coherence times in quantum system design for quantum information science and sensing applications.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
NASA NTRS2005

Origin of the Eumetazoa: testing ecological predictions of molecular clocks against the Proterozoic fossil record

Molecular clocks have the potential to shed light on the timing of early metazoan divergences, but differing algorithms and calibration points yield conspicuously discordant results. We argue here that competing molecular clock hypotheses should be testable in the fossil record, on the principle that fundamentally new grades of animal organization will have ecosystem-wide impacts. Using a set of seven nuclear-encoded protein sequences, we demonstrate the paraphyly of Porifera and calculate sponge/eumetazoan and cnidarian/bilaterian divergence times by using both distance [minimum evolution (ME)] and maximum likelihood (ML) molecular clocks; ME brackets the appearance of Eumetazoa between 634 and 604 Ma, whereas ML suggests it was between 867 and 748 Ma. Significantly, the ME, but not the ML, estimate is coincident with a major regime change in the Proterozoic acritarch record, including: (i) disappearance of low-diversity, evolutionarily static, pre-Ediacaran acanthomorphs; (ii) radiation of the high-diversity, short-lived Doushantuo-Pertatataka microbiota; and (iii) an order-of-magnitude increase in evolutionary turnover rate. We interpret this turnover as a consequence of the novel ecological challenges accompanying the evolution of the eumetazoan nervous system and gut. Thus, the more readily preserved microfossil record provides positive evidence for the absence of pre-Ediacaran eumetazoans and strongly supports the veracity, and therefore more general application, of the ME molecular clock.

NASA Discipline Evolutionary Biology↗
DOE OSTI2024

Influence of pump laser fluence on ultrafast myoglobin structural dynamics

High-intensity femtosecond pulses from an X-ray free-electron laser enable pump–probe experiments for the investigation of electronic and nuclear changes during light-induced reactions. On timescales ranging from femtoseconds to milliseconds and for a variety of biological systems, time-resolved serial femtosecond crystallography (TR-SFX) has provided detailed structural data for light-induced isomerization, breakage or formation of chemical bonds and electron transfer. However, all ultrafast TR-SFX studies to date have employed such high pump laser energies that nominally several photons were absorbed per chromophore. As multiphoton absorption may force the protein response into non-physiological pathways, it is of great concern whether this experimental approach allows valid conclusions to be drawn vis-à-vis biologically relevant single-photon-induced reactions. Here we describe ultrafast pump–probe SFX experiments on the photodissociation of carboxymyoglobin, showing that different pump laser fluences yield markedly different results. In particular, the dynamics of structural changes and observed indicators of the mechanistically important coherent oscillations of the Fe–CO bond distance (predicted by recent quantum wavepacket dynamics) are seen to depend strongly on pump laser energy, in line with quantum chemical analysis. Our results confirm both the feasibility and necessity of performing ultrafast TR-SFX pump–probe experiments in the linear photoexcitation regime. We consider this to be a starting point for reassessing both the design and the interpretation of ultrafast TR-SFX pump–probe experiments such that mechanistically relevant insight emerges.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Control over Conformational Landscapes of Polypeptoids by Monomer Sequence Patterning

The ability to program chain conformation and structure through control over the monomer sequence of synthetic polymers has broad implications for next-generation material design. While related problems of protein-folding and de novo design have generated accurate predictions of 3D folded chain structures, generalization to synthetic polymers remains intractable due to the requirement of large structural databases and the intrinsically disordered nature of polymer building blocks. In this work, polypeptoids, a class of peptidomimetic synthetic polymers, are utilized to build a general workflow for the study of relationships between monomer sequence and dynamic 3D chain structure in solution. Furthermore, this work demonstrates how control over the monomer sequence can alter the conformational landscape of synthetic polymers to deviate dramatically from classical chain statistics. Specifically, the distribution of end-to-end distances, as measured by double electron-electron resonance spectroscopy in dilute solvent, is systematically skewed towards shorter distances with an increasing number of hydrophobes and further refined by hydrophobe arrangement in amphiphilic polypeptoid chains.

36 MATERIALS SCIENCE↗
DOE OSTI2023

HLA-Clus: HLA class I clustering based on 3D structure

In a previous paper, we classified populated HLA class I alleles into supertypes and subtypes based on the similarity of 3D landscape of peptide binding grooves, using newly defined structure distance metric and hierarchical clustering approach. Compared to other approaches, our method achieves higher correlation with peptide binding specificity, intra-cluster similarity (cohesion), and robustness. Here we introduce HLA-Clus, a Python package for clustering HLA Class I alleles using the method we developed recently and describe additional features including a new nearest neighbor clustering method that facilitates clustering based on user-defined criteria. The HLA-Clus pipeline includes three stages: First, HLA Class I structural models are coarse grained and transformed into clouds of labeled points. Second, similarities between alleles are determined using a newly defined structure distance metric that accounts for spatial and physicochemical similarities. Finally, alleles are clustered via hierarchical or nearest-neighbor approaches. We also interfaced HLA-Clus with the peptide:HLA affinity predictor MHCnuggets. By using the nearest neighbor clustering method to select optimal allele-specific deep learning models in MHCnuggets, the average accuracy of peptide binding prediction of rare alleles was improved. The HLA-Clus package offers a solution for characterizing the peptide binding specificities of a large number of HLA alleles. This method can be applied in HLA functional studies, such as the development of peptide affinity predictors, disease association studies, and HLA matching for grafting. HLA-Clus is freely available at our GitHub repository (https://github.com/yshen25/HLA-Clus).

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

Blocking C-terminal processing of KRAS4b via a direct covalent attack on the CaaX-box cysteine

RAS is the most frequently mutated oncogene in cancer. RAS proteins show high sequence similarities in their G-domains but are significantly different in their C-terminal hypervariable regions (HVR). These regions interact with the cell membrane via lipid anchors that result from posttranslational modifications (PTM) of cysteine residues. KRAS4b is unique as it has only one cysteine that undergoes PTM, C185. Small molecule covalent modification of C185 would block any form of prenylation and subsequently inhibit attachment of KRAS4b to the cell membrane, blocking its biological activity. We translated this concept to the discovery and development of disulfide tethering screen hits into irreversible covalent modifiers of C185. These compounds inhibited proliferation of KRAS4b-driven mouse embryonic fibroblasts, but not cells driven by N-myristoylated KRAS4b that harbor a C185S mutation and are not dependent on C185 prenylation. Top–down proteomics was used to confirm target engagement in cells. These compounds bind in a pocket formed when the HVR folds back between helix 3 and 4 in the G-domain (HVR-α3-α4). This interaction can happen in the absence of small molecules as predicted by molecular dynamics simulations and is stabilized in the presence of C185 binders as confirmed by small-angle X-ray scattering and solution NMR. NOESY-HSQC, an NMR approach that measures internuclear distances of 6 Å or less, and structure analysis identified the critical residues and interactions that define the HVR-α3-α4 pocket. Further development of compounds that bind to this pocket could be the basis of a new approach to targeting KRAS cancers.

C185↗
DOE OSTI2024

Insights into the action of phylogenetically diverse microbial expansins on the structure of cellulose microfibrils

Microbial expansins (EXLXs) are non-lytic proteins homologous to plant expansins involved in plant cell wall formation. Due to their non-lytic cell wall loosening properties and potential to disaggregate cellulosic structures, there is considerable interest in exploring the ability of microbial expansins (EXLX) to assist the processing of cellulosic biomass for broader biotechnological applications. Herein, EXLXs with different modular structure and from diverse phylogenetic origin were compared in terms of ability to bind cellulosic, xylosic, and chitinous substrates, to structurally modify cellulosic fibrils, and to boost enzymatic deconstruction of hardwood pulp. Five heterogeneously produced EXLXs (Clavibacter michiganensis; CmiEXLX2, Dickeya aquatica; DaqEXLX1, Xanthomonas sacchari; XsaEXLX1, Nothophytophthora sp.; NspEXLX1 and Phytophthora cactorum; PcaEXLX1) were shown to bind xylan and hardwood pulp at pH 5.5 and CmiEXLX2 (harboring a family-2 carbohydrate-binding module) also bound well to crystalline cellulose. Small-angle X-ray scattering revealed a 20–25% increase in interfibrillar distance between neighboring cellulose microfibrils following treatment with CmiEXLX2, DaqEXLX1, or NspEXLX1. Correspondingly, combining xylanase with CmiEXLX2 and DaqEXLX1 increased product yield from hardwood pulp by ~ 25%, while supplementing the TrAA9A LPMO from Trichoderma reesei with CmiEXLX2, DaqEXLX1, and NspEXLX1 increased total product yield by over 35%. This direct comparison of diverse EXLXs revealed consistent impacts on interfibrillar spacing of cellulose microfibers and performance of carbohydrate-active enzymes predicted to act on fiber surfaces. These findings uncover new possibilities to employ EXLXs in the creation of value-added materials from cellulosic biomass.

09 BIOMASS FUELS↗
DOE OSTI2024

Tertiary folds of the SL5 RNA from the 5 ' proximal region of SARS-CoV-2 and related coronaviruses

Coronavirus genomes sequester their start codons within stem-loop 5 (SL5), a structured, 5' genomic RNA element. In most alpha- and betacoronaviruses, the secondary structure of SL5 is predicted to contain a four-way junction of helical stems, some of which are capped with UUYYGU hexaloops. Here, using cryogenic electron microscopy (cryo-EM) and computational modeling with biochemically determined secondary structures, we present three-dimensional structures of SL5 from six coronaviruses. The SL5 domain of betacoronavirus severe-acute-respiratory-syndrome-related coronavirus 2 (SARS-CoV-2), resolved at 4.7 Å resolution, exhibits a T-shaped structure, with its UUYYGU hexaloops at opposing ends of a coaxial stack, the T’s “arms.” Further analysis of SL5 domains from SARS-CoV-1 and MERS (7.1 and 6.4 to 6.9 Å resolution, respectively) indicate that the junction geometry and inter-hexaloop distances are conserved features across these human-infecting betacoronaviruses. The MERS SL5 domain displays an additional tertiary interaction, which is also observed in the non-human-infecting betacoronavirus BtCoV-HKU5 (5.9 to 8.0 Å resolution). SL5s from human-infecting alphacoronaviruses, HCoV-229E and HCoV-NL63 (6.5 and 8.4 to 9.0 Å resolution, respectively), exhibit the same coaxial stacks, including the UUYYGU-capped arms, but with a phylogenetically distinct crossing angle, an X-shape. As such, all SL5 domains studied herein fold into stable tertiary structures with cross-genus similarities and notable differences, with implications for potential protein-binding modes and therapeutic targets.

59 BASIC BIOLOGICAL SCIENCES↗